Verified Job
Engineering - Architecture
AI Engineer
Engineering - Architecture
#729943
Remote / WFH
Tesselonix Private Limited
Job Description
We are looking for an AI Engineer to build, train, and optimize computer vision systems for real-time highway traffic analysis. The core focus of this role is improving accuracy and reliability of image and video based detection systems, including ANPR (automatic number plate recognition) and traffic compliance classification (seatbelt detection, helmet detection, vehicle categorization, and similar use cases).
Key Responsibilities
• Design and scale pipelines to process continuous video feeds and high resolution images from highway camera infrastructure
• Improve ANPR reading accuracy across varying lighting, weather, angle, and speed conditions
• Train, fine tune, and optimize deep learning models for traffic compliance classification: seatbelt detection, helmet detection, multi class vehicle categorization, and related tasks
• Continuously benchmark and improve model accuracy through better data curation, augmentation, and retraining cycles
• Optimize models for high throughput, low latency inference on edge computing hardware
• Curate, label, and manage large scale image/video datasets for ongoing model improvement
Required Skills & Experience
• Minimum 3 years of hands on industry experience specifically in image processing and video processing (not just general ML)
• Proven experience improving accuracy of classification/detection models in production, not just training on benchmark datasets
• Strong proficiency in PyTorch or TensorFlow, and OpenCV
• Solid experience with object detection (YOLO variants), image classification, and object tracking
• Experience working with real world outdoor/dynamic video data (traffic, surveillance, or similar domains preferred)
• Strong Python skills, production ready code
Good to Have
• Prior experience specifically with ANPR systems
• Experience with RTSP/GStreamer or real time video streaming pipelines
• Model optimization for edge deployment (TensorRT, ONNX, quantization)
• Experience with LLMs/VLMs for multimodal reasoning (not mandatory)
Key Responsibilities
• Design and scale pipelines to process continuous video feeds and high resolution images from highway camera infrastructure
• Improve ANPR reading accuracy across varying lighting, weather, angle, and speed conditions
• Train, fine tune, and optimize deep learning models for traffic compliance classification: seatbelt detection, helmet detection, multi class vehicle categorization, and related tasks
• Continuously benchmark and improve model accuracy through better data curation, augmentation, and retraining cycles
• Optimize models for high throughput, low latency inference on edge computing hardware
• Curate, label, and manage large scale image/video datasets for ongoing model improvement
Required Skills & Experience
• Minimum 3 years of hands on industry experience specifically in image processing and video processing (not just general ML)
• Proven experience improving accuracy of classification/detection models in production, not just training on benchmark datasets
• Strong proficiency in PyTorch or TensorFlow, and OpenCV
• Solid experience with object detection (YOLO variants), image classification, and object tracking
• Experience working with real world outdoor/dynamic video data (traffic, surveillance, or similar domains preferred)
• Strong Python skills, production ready code
Good to Have
• Prior experience specifically with ANPR systems
• Experience with RTSP/GStreamer or real time video streaming pipelines
• Model optimization for edge deployment (TensorRT, ONNX, quantization)
• Experience with LLMs/VLMs for multimodal reasoning (not mandatory)
